1916] ON THE DISCHARGE OF GASES UNDER HIGH PRESSURES 415 On the experimental side it would be of importance to examine, with more accuracy than has hitherto been attained, whether the asserted independence of the discharge of the pressure in the receiving vessel (supposed to be less than a certain fraction of that in the discharging vessel) is absolute, and if not to ascertain the precise law of departure. To this end it would seem necessary to abandon the method followed by more recent workers in which compressed gas discharges into the open, and to fall back upon the method of Saint-Venant and Wantzel where the discharge is from atmospheric pressure to a lower pressure. The question is whether any alteration of discharge is caused by a reduction of this lower pressure beyond a certain point. To carry out ..the investigation on a sufficient scale would need a powerful air-pump capable of absorbing the discharge, but otherwise the necessary apparatus is simple. In order to measure the discharge, or at any rate to determine whether it varies or not, the passage of atmospheric air to the nozzle might be somewhat choked. The accompanying diagram will explain the idea. A is the nozzle, which would be varied in different series of experiments; B the recipient, partially exhausted, vessel.; C the passage bo the air-pump. Above the nozzle is provided a closed chamber E D into which the external air has access through a metal gauze D, and where consequently the pressure is a little below atmospheric. F represents (dia-grammatically) a pressure-gauge, or micromanometer, whose reading would be constant as long as the discharge remains so. Possibly an aneroid barometer would suffice; in any case there is no difficulty in securing the necessary delicacy*. Another manometer of longer range, but only ordinary sensitiveness, would register the low pressure in B. In this way there should be no difficulty in attaining satisfactory results. If F remains unaffected, notwithstanding large alterations of pressure in B, there are no complications to confuse the interpretation. * See for example Phil. Trans, oxovi. A, p. 205 (1901) ; Scientific Papers, Vol. iv. p. 510. [1918. The experiments here proposed have been skilfully carried into effect by Hartshorn, working in my son's laboratory, Proc. Roy. Soo. A, Vol. xoiv. p. 155, 1917.]f the assumption that viscosity may be neglected when a jet moves with high velocity through quiescent air.